“Nothing Is Shown”: a 'Resolute' Response

Philosophical Investigations 26:3 July 2003 ISSN 0190-0536

“Nothing is Shown”: A ‘Resolute’ Response to Mounce, Emiliani, Koethe and Vilhauer

Rupert Read and Rob Deans, University of East Anglia

Part 1: On Mounce on Wittgenstein (Early and Late) on ‘Saying and Showing’

H. O. Mounce published in this journal two years ago now a Criti- cal Notice of the The New Wittgenstein,1 an anthology (edited by Alice Crary and Rupert Read) which is evenly divided between work on Wittgenstein’s early and later writings. The bulk of Mounce’s article was devoted to those contributions primarily concerned with the Tractatus.2 There is a straightforward sense in which this selective focus is natural. The pertinent contributions – most conspicuously those by Cora Diamond and James Conant – describe a strikingly un- orthodox interpretation of Wittgenstein’s early book on which it is depicted as having an anti-metaphysical aim. Mounce takes an interest in this interpretation because he believes that, in characterizing the Tractatus in anti-metaphysical terms, it misrepresents the central Tractarian doctrine of ‘saying and showing’ – a doctrine which he understands in terms of the idea that “metaphysical truths, though they cannot be stated, may nevertheless be shown”(186). Mounce argues that Diamond and Conant et al. fail to treat this doctrine as “one that Wittgenstein himself advances,” and he claims that they therefore make Wittgenstein’s thought “less original than one might otherwise suppose” (186) by implying that it is “indistinguishable from positivism” in the sense of “not even attempt[ing] to provide positive knowledge [and] conﬁn[ing] itself to removing the confu-

1. A. Crary and R. Read, eds., Routledge, London, 2000. See Mounce, “Critical Notice of The New Wittgenstein,” Philosophical Investigations,vol. 24, no. 2, 2001, pp. 185–192. All references in Part 1 of the text are to this article, unless otherwise stated. 2. See Part II of The New Wittgenstein (op. cit., note 1).The essays in Part I, in contrast, are primarily concerned with Wittgenstein’s later thought.

© Blackwell Publishing Ltd. 2003, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. 240 Philosophical Investigations sions which have been inflicted on us by traditional philosophy or metaphysics” (187). Moreover, although Mounce has relatively little to say about Wittgenstein’s later thought, he adds that he thinks the Diamond-Conant interpretation prevents us from appreciating the way in which the say-show distinction continues to figure in Wittgenstein’s post-Tractatus writings (191–192). Mounce’s criticism of the Diamond-Conant interpretation of the Tractatus is clear and forcefully argued. It also derives a certain inde- pendent interest from its resemblance to other emerging criticisms – e.g., as Mounce himself points out (185–186), the one that P. M. S. Hacker’s develops in an appendix to The New Wittgenstein.3 Nevertheless, what we hope to demonstrate here is that, despite its initial appeal, the criticism fundamentally misrepresents some of the most basic claims of the Diamond-Conant interpretation, including its claims about the say-show distinction, and, further, that these mis- representations have a significance that extends beyond the Tractatus. One of Diamond’s and Conant’s goals in trying to isolate an anti- metaphysical strain of thought often overlooked in the Tractatus is to demonstrate the persistence (and development) of the same strain in Wittgenstein’s later writings. It follows that, to the extent that Mounce misdescribes the main concerns of their interpretation, he also inevitably fails to capture the kind of continuity in Wittgenstein’s thought that preoccupies them. A helpful, preliminary way to characterize what is distinctive about the Diamond-Conant interpretation of the Tractatus is to note that its proponents read Wittgenstein as accenting the idea – one they often represent him as inheriting from Frege – of the ‘primacy of the proposition’. The author of the Tractatus, as Diamond and Conant understand him, takes the proposition, or complete thought, to be the sole proper object of logical analysis.4 Thus, as they read him,Wittgenstein (to put it in terms Frege uses to describe his own philosophical method) does not “begin with concepts and put them together to form a thought or judgment” but rather “come[s] by the parts of a thought by analysing the thought.”5 To say that Wittgen- stein’s route to the parts of a thought is invariably through analysis

3. Ibid., chapter 14. 4. Wittgenstein of course progressively broadens Frege’s more tentative ‘contextualism’. See e.g. Read on this, on pp. 76–7 of The New Wittgenstein. 5. Gottlob Frege, Posthumous Writings,H.Hermes et al., eds., P.Long and R.White, trans., Chicago, University of Chicago Press, 1979, p. 253.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 241 of the whole is to imply that he thinks that our ability to identify the logical roles played by the signs that compose a sentence depends on our understanding of the thought the sentence expresses and, further, that there can therefore be no such thing as identifying the logical roles played by the parts of a nonsensical sentence.This means that Wittgenstein, as proponents of the Diamond-Conant interpretation read him, repudiates the idea that some bits of nonsense are logically distinct from mere gibberish6 in the sense of being pro- duced by clashes between logically incompatible categories, or ‘violations of logical syntax’. Now, Diamond and Conant et al. frequently describe this idea – viz., the idea that some sentences are nonsense on account of the particular illegitimate kind of thing their supposedly logically incompatible parts attempt (but fail) to say – as the hallmark of a substantial conception of nonsense. And one of their chief contentions is that in the Tractatus Wittgenstein endorses a very different (and, in their terminology) austere conception of nonsense on which – however psychologically plausible we may ﬁnd the thought that we are sometimes confronted with category clashes that differ from mere gibberish – all bits of nonsense are nevertheless logically indistinguishable.7 Indeed, these commentators on the Tractatus maintain that if we are to read Wittgenstein’s book with understanding we need to see that this austere conception underwrites even his famous closing claim (i.e., Tractatus 6.54) that sentences of his book are nonsense. This brings us to our main objection to Mounce’s account of the Diamond-Conant interpretation. It fails to register the importance proponents of the ‘New Wittgenstein’ interpretation place on this distinction between austere and substantial conceptions of nonsense. Mounce simply situates his discussion of different interpretations of the Tractatus within a space of possibilities determined by (what Diamond and Conant et al. call) the substantial conception. He assumes that Wittgenstein thinks that metaphysical sentences are nonsense on account of the kind of substantial thing they attempt –

6. Such as, e.g., ‘ab sur ah’ or ‘higgly piggly wiggly’. 7. For references to treatments of the distinction between ‘austere’ and ‘substantial’ conceptions of nonsense within The New Wittgenstein (op. cit., note 1), see esp. Diamond,“Ethics, Imagination and the Tractatus,” pp. 153 and 165 and Conant,“Elu- ciation and Nonsense in Frege and Early Wittgenstein,” pp. 176–179, 185, 188–189, 191 and 195–196. See also below.

© Blackwell Publishing Ltd. 2003 242 Philosophical Investigations unsuccessfully – to say. And, on the basis of this assumption, he declares that a central exegetical issue confronting commentators on the Tractatus concerns whether or not these (substantially) nonsensical, metaphysical sentences are capable of ‘showing’ us things about the nature of thought and language that cannot properly be put into words. He thus arrives at the view that the only alternative to the kind of ‘ineffability’ interpretation he himself favours – on which Wittgenstein is taken to hold that metaphysical sentences illuminate ineffable truths – is a ‘positivistic’ interpretation on which such sentences are excluded from playing any illuminating role. And, since he recognizes that Diamond and Conant et al. are hostile to ineffability readings, he concludes that they must be fans of positivistic ones. But this conclusion is plainly false. The trouble is that, in drawing it, Mounce in effect suggests that Diamond and Conant self- consciously attribute to Wittgenstein the very – substantial – conception of nonsense that they in fact take him to reject.8 Diamond and Conant read Wittgenstein as distancing himself from the substantial conception of nonsense, so they reject both Mounce’s assumption that Wittgenstein conceives metaphysical sentences as substantially nonsensical and his assumption that Wittgenstein is concerned with delivering an answer to a question about whether metaphysical sentences, thus conceived, ‘show’ us things about the logical structure of language. Proponents of the Diamond-Conant interpretation are concerned to record the fact that, within the Tractatus, the notion of showing [zeigen] is never used in reference to nonsense [Unsinn] but only in reference to legitimate, well-formed propositions.9 Here we might recall two central Tractarian remarks on showing: Propositions cannot represent logical form: it is mirrored in them. What mirrors itself in language, language cannot represent. What expresses itself in language, we cannot express by language. Propositions show the logical form of reality. They display it (4.121). What can be shown cannot be said (4.1212).10

8. Moreover, Conant himself argues exactly this case in some detail in his paper in The New Wittgenstein. It is surprising that Mounce missed this. 9. On this point, see esp. Conant, “Elucidation and Nonsense . . .” in The New Wittgenstein (op. cit., note 1), pp. 178–179 and notes 11 and 20. 10. Here we draw on, and revise, both the Ogden and Pears & McGuinness trans- lations of the Tractatus.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 243 When proponents of the Diamond-Conant interpretation discuss the Tractatus say-show distinction, it is this distinction they have in mind.11 Further, although they suggest that Wittgenstein ultimately wants to get us to relinquish the idea of ‘what can be shown but not said’ as plainly nonsensical, they also believe that he takes the process through which we become dissatisfied with it, if we do in fact become dissatisfied, to be one that is simultaneously morally and intellectually demanding. There is in this respect an immense philosophical gulf between Wittgenstein’s thought, as Diamond and Conant et al. understand it, and positivistic modes of thought.12 We might sketch Diamond-and-Conant’s basic account of the Tractarian attitude towards saying and showing as follows. These philosophers read early sections of the Tractatus as acknowledging that, at moments at which we are struck by the mere fact of our ability to think and talk about the world, we may think that metaphysics can explain this ability, and we may accordingly try to for- mulate metaphysical sentences that express what (we imagine) our propositions show about the ‘logical form of reality’. Further, they read the Tractatus as a whole as trying to teach us that when we take ourselves to be offering such metaphysical explanations we are confused about what we are doing. We are using signs to which we ourselves have assigned no meaning (Tractatus, 6.53) and putting ourselves in a position in which – as a patient interlocutor might get us to recognize – no intelligible use of the relevant signs cap- tures what we confusedly imagine we want to say. This, according to Diamond and Conant et al., is the route which the Tractatus takes to dismantling its own say-show distinction. Its aim is to bring us to the ‘austere’ recognition that our allegedly metaphysical utterances are empty and that we ourselves have no use for the idea, once dear to us, of ‘something that can be shown and not said’. The Tractar- ian distinction between saying and showing thus turns out to figure as an important preliminary stage on what might be described as the reader’s journey of self-discovery.13

11. As Conant makes abundantly clear, e.g. in n. 19 of his “Elucidation and nonsense” paper. 12. These intellectual demands lead to one’s knowing one’s way about; that is, ulitmately, to a kind of ‘self-knowledge’. Not to items of (ineffable) knowledge. 13. We have just given a brief sketch of what Diamond and Conant et al. see as a central anti-metaphysical strand of thought in the Tractatus. It may worth emphasiz- ing that our willingness to agree with them that this strand of thought is in the Trac-

© Blackwell Publishing Ltd. 2003 244 Philosophical Investigations The break here with positivism is decisive. The classical positivist holds that we are prevented by the very structure of language from entering into metaphysical explanations. She is committed to the view that there is something at least barely coherent which the nature of language forever cuts us off from doing.14 In contrast, the austere thinker that Diamond and Conant believe the Tractatus envisions as its ideal reader rejects the very idea of limits of language that pro- hibit us from saying certain substantial kinds of things.15 This thinker holds that, to the extent that it is meaningful to talk about limits of thought or language at all, such limits must be drawn, as Wittgen- stein puts it in the Preface to the Tractatus, in language – i.e., in our humble and ordinary efforts to make sense of particular combinations of signs. And, indeed, it is in virtue of the fact that the Tracta- tus, as Diamond and Conant et al. read it, thus aims to transform its reader’s conception of what limits of language are like – and, more speciﬁcally, to show its reader that there are no such things as limits that can somehow relieve us of the need to clarify our own thinking – that they think it is properly understood as a contribution to ethics.16 Let us turn now brieﬂy to Mounce’s suggestion that the Diamond-Conant interpretation fosters misunderstanding of the sense in which a say-show distinction is projected into Wittgenstein’s tatus does not necessarily commit us (still less them!) – as Mounce seems to think – to claiming “that the Tractatus is [not] even covertly metaphysical” (186) (See Parts 2 and 3 below, for our discussion of the different respects in which ‘New Wittgen- steinians’ can take the Tractatus to be metaphysical). It is not inconsistent to endorse basic tenets of the Diamond-Conant interpretation while also holding, as arguably they themselves do, that (e.g.) the Tractatus view of the kind of regularity constitu- tive of “logical space” encodes a metaphysical perspective that, in his later treatment of rule-following, Wittgenstein attempts to dismantle. Relevant issues come up in most of the essays on the Tractatus in The New Wittgenstein (op. cit., note 1). See esp. Diamond, “Does Bismarck Have a Beetle in his Box?” 14. In more general terms, this is the false lure of ‘Anti-Realism’. See Read’s “Is “What is time?” a good question to ask?”, Philosophy 77 (2002), pp. 193–209, his “Time to stop trying to provide an account of time”, forthcoming in Philosophy 78 (2003); and his “What does “signify” signify?”, Philosophical Psychology 14:4 (2001), pp. 499–514. 15. For helpful discussion of how the author of the Tractatus conceives of limits of language, see esp. Diamond, “Ethics, Imagination,” The New Wittgenstein,op. cit., note 1, pp. 150ff. See also Read’s “What does “signify” signify?” (op.cit.). 16. In this respect, the Diamond-Conant interpretation provides a compelling account of Wittgenstein’s famous claim that his early book has an ethical point. In this connection, see also recent work on the Tractatus by Michael Kremer and Eli Friedlander.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 245 later writings. Consider, to begin with, Mounce’s own view of how the distinction gets carried over. Towards the end of his Critical Notice, Mounce points out that the later Wittgenstein is receptive to the idea of many different kinds of uses of language and that he treats sentences that describe ways in which we use linguistic signs as expressing legitimate – grammatical – propositions (191). Mounce claims that such grammatical propositions have the role in Wittgen- stein’s later writings – it is same the role that Mounce thinks the Tractatus assigns to certain nonsensical metaphysical sentences – of ‘showing’ the logical structure of language. He writes that:

grammatical propositions are entirely parasitic on what shows itself in language; their function, indeed, is to draw our attention to what shows itself there. In effect we have the same distinction between saying and showing (192).

Here Mounce is operating with a quite standard picture of the development of Wittgenstein’s thought. He is taking it for granted that Wittgenstein moves from claiming, early on, that metaphysical sentences determine a limit separating the kinds of things that can be said from the (substantial) kinds of things that cannot to later rejecting this view and claiming that grammar determines a limit separating the kinds of things that can be said from the (substantial) kinds of things that cannot. Mounce starts from this standard picture and then suggests that the question of whether or not the later Wittgenstein retains (what Mounce understands as) his early say-show distinction depends on how he takes grammar to play the relevant determinative role. According to Mounce, there are two alternatives. Either we read Wittgenstein – as Mounce recommends – as claiming that grammatical propositions reveal, or ‘show’, deep necessary truths about the nature of language, or we read him as claiming that grammatical propositions describe mere contingencies about our discursive practices. Since Mounce recognizes that philosophers sympathetic to the Diamond-Conant interpretation of the Tractatus are hostile to the former ‘showing’ approach to Wittgenstein’s later writings, he assumes that they understand Wittgenstein’s interest in grammar as an interest in mere contingencies and that they therefore conceive Wittgenstein as a thorough-going conventionalist – as a thinker who holds “that language is explicable in terms of convention” (192). This assumption is problematic because, as we saw above, Mounce understands grammar an arbiter of substantial nonsense. Mounce’s

© Blackwell Publishing Ltd. 2003 246 Philosophical Investigations discussion of Wittgenstein’s later work, like his discussion of the Tractatus, takes place in a space of alternatives determined by the substantial conception of nonsense. So Mounce is, once again, in the position of suggesting that Diamond and Conant are happy to ascribe to Wittgenstein the very conception of nonsense they think he rejects throughout his work.17 By contrast, their thought is in fact that, when, in his later writings, Wittgenstein urges us to look at ways in which we use words that interest us in philosophy, he is reminding us that there is no method of drawing limits to language that somehow trumps, or supersedes, our everyday efforts to make sense of utterances and inscriptions. They thus read Wittgenstein’s later writings as having the same basic ethical orientation that they take to be characteristic of the Tractatus.Wittgenstein’s aim in both cases is to remind us that we cannot avoid responsibility for the accounts we give of where the limits of thought and language, the limits of our ability to make sense of ourselves and each other, are properly drawn.

Part 2: Advancing the Resolute Agenda: ‘Weak’ and ‘Strong’ Resolutisms on Effability

A useful way of putting part what emerges from the argument of Part 1, above, is that, ironically, it is the authors featured in the New Wittgenstein who in the end perhaps have the most plausible claim to be true defenders of the idea of saying and showing as something

17. It is perhaps worth stressing that our point here is not that there are no commentators who read the later Wittgenstein as a conventionalist of the sort Mounce has in mind. The later Wittgenstein is standardly read as the relevant sort of conventionalist. Moreover, although we cannot discuss this issue here, we are inclined to think that Mounce is right to suspect that P.M.S. Hacker is properly understood as developing a version of such a standard reading (191–192).The point we are making is, not that conventionalist readings are out of favour, but rather that Mounce’s assumption that the Diamond-Conant interpretation of the Tractatus has a tendency to push us towards such readings of Wittgenstein’s later work is based on a fundamental misunderstanding. One should bear in mind in this connection that Wittgenstein thought that Carnap (who was, not incidentally, generally a conventionalist) had utterly missed his “fundamental” thinking, in missing the point of the closing paragraphs of the Tractatus (see e.g. n.3 of Conant’s “Elucidation and nonsense in Frege and early Wittgenstein”, in The New Wittgenstein). One can hardly help thinking here of Tractatus 4.0312, too.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 247 important in Wittgenstein. Because only ‘we’ both make sense of what Wittgenstein actually says about saying and showing (i.e., about senseless satze showing that they say nothing, and nonsenses not saying or showing anything), and yet allow that there can be some process of attaining what we might risk calling insight (the kind of insight involved in understanding oneself ) as a result of reading Wittgenstein’s early work. We ‘New Wittgensteinians’ tend to stress the issue of sense and nonsense so much because it has been neglected – but we also have something important to say about showing and about (what others misleadingly call) ‘showing’ (i.e., about the attainment of some knowing of one’s way about, through the learning about oneself that happens when one engages in the therapeutic activity that Wittgenstein urges, in the Tractatus and the Investigations alike). This cannot be said of Mounce, or of Hacker, or even of Lynette Reid. But the ‘New Wittgensteinians’ do not speak with the narrowness of one voice on these matters. The revisionary reading of the Trac- tatus (in the context of the entirety of Wittgenstein’s corpus) associ- ated most especially with the pioneering work of Cora Diamond and its subsequent development by James Conant is a fairly loose programme for action, which so far has only been partially and variously carried out by those who find in the writings of Diamond and Conant both insight and inspiration for carrying this work forward. The writings of Diamond and Conant themselves are self- consciously a prolegomena for future work. This ought to sound a note of caution to those critics who want to speak of ‘resolutism’,18 or ‘the resolute reading’, as expressing some completed or even fully defined project. As we urged in Part 1, above, one thing that ‘resolutism’ is not, despite its sometime concentration on the Tractatus, is a belief that the best of Wittgenstein’s philosophy is to be found in the early work. Rather, what actually motivates most ‘resolutists’ is Wittgen- stein’s later work, for it is this that is thought by ‘resolutists’ to be more philosophically interesting and more relevant to contemporary philosophical discussions. The significance of the Tractatus is that it refocuses attention on just what were the continuities and disconti-

18. We mention this label, for brevity’s sake, with doubtful feelings: it MUST NOT be read as committing us (or Wittgenstein) to ‘a philosophical position’ or any such like.

© Blackwell Publishing Ltd. 2003 248 Philosophical Investigations nuities between the early and later works and, in so doing, not only calls into question the established readings of this work, but more importantly the established readings of the later work too. As ‘resolutism’ is not a single viewpoint it is useful to distinguish between at least two directions that the reading-project seems to be currently heading in.These could be called the strong version and the weak version. No evaluative judgment is intended in using these two designations; they merely express an important difference in emphasis as to how the resolute reading appears to be developing. As a reading of the Tractatus,‘resolutism’ can (as explained above) be thought of as embracing the following two core commitments: an austere conception of nonsense; and the rejection of ‘positivism’ (or, more broadly, ‘Carnapianism’) and ‘ineffabilism’. ‘Resolutism’ maintains that nonsense is only ever to be ‘understood’ via an austere as opposed to a substantial conception of nonsense. As discussed in Part 1, this commitment to an austere conception of nonsense follows from the contextualism that resolutism believes to be at the heart of Wittgenstein’s discussion of the distinction between signs and symbols in the Tractatus, and the segmentation of significant propositions into their meaning-bearing constituents. The Tractatus makes a distinction between propositional signs (Satzzeichen) that are sensical (sinnvoll), those that are senseless (sinnlos) and those that are nonsense (unsinnig). Into the first category fall the propositions of science; these say how things stand in the world and, if true, show how things stand. Into the second category fall ‘the propositions of logic’; these say nothing about how things stand in the world and show that they say nothing. Into the third category fall all other ‘propositions’ (sic.); these say and show nothing at all. It is only the propositions of natural science that are sensical and can properly be called significant propositions (Satz), although the term is ambiguously applied throughout the Tractatus (deliberately so, according to ‘resolutism’) to the so-called propositions of logic and to all other ‘propositions’. Into which category any given propositional sign falls is determined by comparing it with other propositional signs of a similar form that express significant propositions. To this end a concept script (Begriffsschrift) or notation can be used to make perspicuous the similarities and dissimilarities in logical form that characterise propositions. This discussion clearly assumes familiarity with the logical systems

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 249 of both Frege and Russell; the Tractatus also seems to be taking for granted the utility and genuineness of the results of philosophical logic. It seems a straightforward question, therefore, to ask to just what extent does Wittgenstein share Frege’s and Russell’s conception of philosophical logic and equally as straightforward to try and answer this question by examining what Wittgenstein actually has to say in the Tractatus concerning the logical systems they devised. The more established readings of the Tractatus proceed precisely on these lines and ﬁnd in the Tractatus points of agreement and disagreement with Frege and Russell and consequently the expression by Wittgen- stein of a number of theoretical commitments. This is a puzzle; for the Preface explicitly states that the Tractatus is not a textbook (Lehrbuch), yet on the other hand there is apparently a technical discussion of the essence of the proposition. Just what is going on? ‘Resolutism’ sees the Tractatus as operating at several different ‘levels’. At one level Wittgenstein can be seen as engaging with the logical systems of Frege and Russell in order to highlight errors in their analysis. At another level Wittgenstein can be seen as providing at least a corrective and possibly an alternative to their logical systems. If these are all that Wittgenstein is thought to be doing in the Tractatus, then it does have the appearance of a textbook, or at least of something which could be harmlessly paraphrased into a textbook-form. However,‘resolutism’ thinks that if the book is taken this way, then the most important levels of all are missed. It is not logical analysis that Wittgenstein objects to in the Tractatus,but the misconception Frege and Russell have of what philosophical logic is and can accomplish: namely that logical analysis is a maximally general science of the laws of truth; that it provides a universally applicable framework within which a determinate set of categorial distinctions can ﬁx the interpretation of the signs in which all thought must be expressed; and so can provide a basis for answer- ing philosophical questions. Wittgenstein uses his own set of categorial distinctions and proposals for a concept-script to show how misleading not only the categorial distinctions of Frege and Russell are, but how misleading any categorial distinctions (including his) can be. The point here is not that we cannot or should not make categorial distinctions, but that we can become seduced by the metaphysics of logical analysis. A key difference between the weak and strong versions of resolutism is that the weak version thinks that Wittgenstein attempted to demonstrate in the Tractatus just what a

© Blackwell Publishing Ltd. 2003 250 Philosophical Investigations deflationary and therapeutic conception of logical analysis amounts to.The strong version, however, thinks that Wittgenstein is concerned there to demonstrate not just that a propositional sign can fail to find a definite use in thinking and therefore be nonsense, but also that there is thinking which cannot be completely reduced to and be fully expressed by propositional signs that answer to the general and universal categorial distinctions of the logical systems devised by Frege and Russell, and even those he develops in the Tractatus. The strong version maintains, therefore, that the Tractatus is a deliberately ‘self-refuting’ attempt to establish that no logical system is powerful enough to fully express in a general way the meaningfulness or the meaninglessness of any possible configuration of signs, and that the attempt to do so, including the attempt of the Tractatus, will inevitably result in nonsense. The levels at which the text operates makes it possible to have a technical and detailed discussion of aspects of philosophical logic that are thought to offer genuine results, whilst also objecting to a particular conception what philosophic logic is and should achieve. Further, in so far as philosophical logic does provide insight, it is not its correctness that Wittgenstein affirms, but its utility. Given this, the challenge for ‘resolutism’ is to explicate how the propositions of the Tractatus can elucidate something whilst being nonsensical. In the Tractatus,Wittgenstein is particularly concerned with propositional signs that appear to have a form similar to other significant propositions and that, very importantly, one takes to be expressing philosophical thoughts, but which on examination are found to be nonsense. These can be referred to as ‘philosophical propositions’, but it needs to be kept in mind, for the reasons given above, that applying the term ‘proposition’ is misleading. One of the principal characteristics of ‘resolutism’, as a reading of the Tractatus, is the emphasis it therefore places upon Wittgenstein’s remarks in 6.54, as has been made clear in detail by Conant.19 When Wittgenstein refers – in his crucial discussions with Ogden on how the Tractatus should be translated – to the object of philosophy as “the logical clarification of thoughts” and its result that “the propositions have become clear”, he is concerned with just how it is that we come to know whether a propositional sign has sense, is senseless or is nonsense. Philosophy, as Wittgenstein conceives it, is an 19. Especially, again, in his “Elucidation and nonsense in Frege and early Wittgenstein”.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 251 activity which can assist us in coming to recognise whether a propositional sign that appears to have sense, that we are inclined to think means something, is or is not a signiﬁcant proposition. Further, it achieves this result not by asserting ‘philosophical propositions’, but rather by giving elucidations. What can be concluded from this is that as a philosophical work the Tractatus consists essentially of elucidations; these elucidations concern “philosophic matters”; they are nonsense; and nonsense is to be understood according to an austere as opposed to a substantial conception. The weak version of resolutism sees this as consistent with a continuing role for logical analysis, but only as deﬂationary and therapeutic. The strong version of ‘resolutism’ accepts this continuing therapeutic role for logical analysis, but thinks that Wittgenstein is much more concerned in the Trac- tatus to demonstrate that thinking cannot be completely reduced to and be fully expressed by any logical system and that the attempt to do this results in nonsense. ‘Resolutism’ therefore lays great stress upon what it perceives to be a series of methodological remarks scattered throughout the Tractatus (and also upon Wittgenstein’s comments to Ogden on its translation). ‘Resolutism’ regards these methodological remarks as constituting a ‘Frame’ to the Tractatus. One difference between the weak and strong versions of resolutism is the extent to which these methodological remarks are privileged. In the weak version, the ‘Frame’ can be held onto whilst the rest of the Tractatus is thrown away. In the strong version, the ‘Frame’ too is seen as yet another expression of the impulse towards metaphysics, and is to be surmounted as well. ‘Resolutism’ therefore makes a provisional distinction between the meaning of the ‘propositions’ of the Tractatus concerning “philosophic matters” (which are nonsensical), and a very special kind of use that they can have. It holds that to make the kind of use of these propositions that Wittgenstein intends, and thereby to understand him and not the ‘propositions’, crucially depends upon the reader coming to recognise that these ‘propositions’ are indeed nonsensical. It is through this recognition that the reader is then able to surmount them and ‘come to see the world rightly’. ‘Resolutism’, as a reading, is the attempt to work through the text of the Tractatus acknowledging and accepting this challenge, for what it is to see the world rightly can only be explicated in so far as the reader is able to recognise the propositions of the Tractatus concerning “philosophic matters” as nonsensical.

© Blackwell Publishing Ltd. 2003 252 Philosophical Investigations In the weak version of ‘resolutism’, the propositions of the Tracta- tus are to be seen as rungs in a ladder that the reader must painstak- ingly climb in order to identify the philosophical concerns that motivate the text and to determine just how and in what way the philosophical propositions that are being expressed result in nonsense. What is then thrown away is the cumulative nonsense arising from this engagement with the text. In the strong version, this is not the end of the matter, for the Tractatus, itself, as a whole, demands to be seen as yet another expression of the impulse towards metaphysics; namely that a complete analysis of logical form is possible, and that logical form thus understood determines the limits of the application of signs.What is then thrown away is not just the cumulative nonsense arising from engaging with particular propositions of the text that deal with particular “philosophic matters”, but the very idea inherent in the text that there are hidden necessities that determine the limits of the use of language. One ﬁnally realizes that nonsense in the Tractatus is ultimately parasitic on a particular conception of sense, a conception that is explicated in terms of logical form, and that if there is no such thing as the ineffability of nonsense, then there is no such thing as the effability of sense either. (Thus, via the transitional route of austerity about nonsense, we begin to be ‘returned’ to the details and differences of and among particular utterances – anticipating the ‘trajectory’ of Wittgenstein’s later thinking, even of his increasingly cautious use of the term “nonsense”.) Conant’s paper in The New Wittgenstein, “Elucidation and nonsense in Frege and early Wittgenstein”20 disposes of the two interpretive strategies that, historically-speaking, have dominated the reception of the Tractatus, namely, broadly ‘positivist’ readings, and ‘ineffabilist’ readings (which have predominated in recent years).21 One way of trying to deﬁne ‘resolutism’ is to say that it does not accept that it is possible to imagine a perspective from where one can survey how language represents the world, either from an ‘inter-

20. This paper is an excerpt from a longer manuscript of Conant’s, “The Method of the Tractatus”, which is published in large part in E. Reck’s book, From Frege to Wittgenstein (Oxford: OUP, 2001). 21. Read has a number of papers on the carrying forward of the positivist versus ineffabilist dispute into the interpretation of Wittgenstein’s later work, most recently “The ﬁrst shall be last and the last shall be ﬁrst. . .”, in D. Moyal-Sharrock and W. Brenner (eds.), Investigating ‘On Certainty’ (forthcoming). Read there suggests that John Koethe is among those giving a (covertly) ineffabilist interpretation of Wittgenstein’s later work.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 253 nal’ (positivist) or an ‘external’ (ineffabilist) point of view. However, defining ‘resolutism’ in these terms is very misleading, for it suggests that ‘resolutism’, in denying the possibility of being able to imagine such a perspective, nonetheless allows that there is perhaps ‘something’ here that cannot be done. However, it is just this metaphysical picture (and expression of human finitude) that ‘resolutism’ thinks that Wittgenstein calls into question already in the Tractatus. What is at issue for ‘resolutism’ is not the truth or falsity of this ‘something’ – that is, whether or not it is possible to imagine such a perspective – but rather the intelligibility of this ‘something’ – that is, whether there could be such a thing as being able to imagine a perspective where one can survey how language represents the world. Whether, that is, we have any use for those words, any use for them that we ourselves will take as amounting to anything. According to ‘resolutism’, positivism and ineffabilism think that they can imagine such a perspective. For ‘resolutism’ there is as yet no thing as such a perspective of the kind that positivism and ineffabilism try to imagine. ‘Resolutism’, as a reading of the Tractatus, therefore rejects both Realism and Anti-Realism. It regards the possibility of an ‘internal’ or an ‘external’ point of view upon how language represents world as equally unintelligible. It sees both as an expression of the impulse towards metaphysics, as symptomatic of the very philosophical illness that the Tractatus is itself the diagnosis and attempted cure of. In 6.54, when Wittgenstein writes, “My propositions are elucidatory in this way: he who understands me finally recognizes them as [nonsense]”, this could be taken positivistically to mean that Wittgen- stein’s propositions are nonsense because they fail to meet criteria for what it is for a proposition to have sense, or it could be taken ineffabilistically to mean that even though Wittgenstein’s propositions fail to meet this criteria for what it is for a proposition to have sense and so are nonsense, yet they still show something about what cannot be said. The more established readings (mis)understand 6.54 in just this way. The weak version of resolutism reads 6.54 as part of the ‘Frame’: Wittgenstein does not say in these methodological remarks either that all the propositions of the Tractatus are nonsense, or that all the propositions of the Tractatus are elucidatory. According to Wittgen- stein, “A philosophical work consists essentially of elucidations”, and the result of such elucidations is “that the propositions have become

© Blackwell Publishing Ltd. 2003 254 Philosophical Investigations clear”. The ‘propositions’ referred to in 6.54, if we pay attention to Wittgenstein’s comments to Ogden, concern “philosophic matters”. These are elucidated by coming to realise that certain propositional signs that one takes to be sensical and to have philosophical significance are, in that context of use, nonsensical and that this, in turn, is caused by the failure to give meaning to certain signs in those propositions.The reader comes to see the world rightly through the unsuccessful attempt to try and make sense of the propositions of the Tractatus that attempt to express the philosophical concerns that apparently motivate the text. This leaves open rather than closes down the possibility that the text can be read dialectically, acknowledging the different levels at which the text operates, and avoids the absurd belief that any line of the Tractatus is to be regarded simply as as nonsensical as any line of Jabberwocky.22 The strong version of resolutism reads 6.54 more ‘self-referentially’. It maintains that Wittgenstein is concerned in the Tractatus to demonstrate that thinking cannot be completely reduced to and be fully expressed by anything we are likely to be content to call a ‘logical system’. What is as a consequence nonsense is the assumption of logical analysis that all thinking, in order to be thought, must be captured by a propositional sign that on examination is found to be either sensical, senseless or nonsense – even when it is understood that this cannot be decided in advance and in general by mere inspection of the signs out of which the propositional sign is constructed, but requires attention to be given to its possible logico-syntactical application. What is also nonsense, therefore, is 6.54, itself, and here, at last,Wittgenstein begins to recover our ordinary non-philosophical ways of speaking. There is nothing wrong in characterising something as nonsense; the error lies in regard- ing the determination of sense and nonsense to be a philosophical problem. The nonsense that pre-eminently (but ‘transitionally’) interests Wittgenstein is, loosely put, philosophical nonsense. It is not that there are different kinds of nonsense, some more signiﬁcant than others, but the Tractatus is concerned with unravel- ling how particular propositional signs that are taken to express sense – to have ‘philosophical signiﬁcance’ – are nonsense. A propositional

22. Assuming for present purposes what might be controversial, namely that Jabberwocky is a good example of (gibberish-ish) nonsense.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 255 sign of itself, of course, says nothing. Application is required before it can express a significant proposition. It is not because ‘philosophical propositions’ lack form that they are nonsense; it is because they have no definite use. It is because of the recognizable form they do actually have that ‘philosophical propositions’ are able to beguile us into thinking that they are sensical. However, it is not until an unsuccessful attempt is actually made to make sense of the propositional sign, to consider its possible logico-syntactical application, that it becomes apparent that there has been a failure to give meaning to certain signs in the proposition. But this cannot be determined by inspection alone. The attempt to find sense requires that the propositional sign be ‘taken’ as sensical. Application tries to find the symbol(s) in the sign(s) based on the logical segmentation of significant propositions, particularly of those that have a similar form.This involves experimenting with logical segmentation and the possible meanings of the signs.The sign can only be judged nonsensical when it is realised that despite these efforts at application no sense at all can yet be made of the propositional sign; that is, that no use can be found for it. It is then nonsense not because of any conflict its meaning-bearing constituents have, for it has no meaning-bearing constituents at all. Confusion creeps in because the attempted logico- syntactical application did involve experimenting with logical segmentation and this required trying to find the symbol in the sign, to identify meaning-bearing constituents. However, once the attempt at application has failed all possibility of there remaining any logical segments or meaning-bearing constituents ends, for these can only be found in significant propositions. One of the uses that Wittgenstein envisages for a concept-script in the Tractatus is to make perspicuous how different attempts at application succeed or fail in making sense. Whilst ‘resolutism’ believes that ‘philosophical propositions’ are nonsensical, ‘resolutism’ does not believe that one can say in advance and in general whether any given propositional sign is a ‘philosophical proposition’ or not. It simply cannot be decided by inspection of the signs out of which it is constructed whether a propositional sign is nonsense or not. ‘Resolutism’ thinks that this preoccupation in the Tractatus with “philosophic matters” is supported by 6.53, where the right method in philosophy is contrasted with Wittgenstein’s own in 6.54.Wittgen- stein writes:

© Blackwell Publishing Ltd. 2003 256 Philosophical Investigations “The right method of philosophy would be this. To say nothing except what can be said, i.e. the propositions of natural science, i.e. something that has nothing to do with philosophy: and then always, when someone else wished to say something metaphysical, to demonstrate to him that he had given no meaning to certain signs in his propositions. This method would be unsatisfy- ing to the other – he would not have the feeling that we were teaching him philosophy – but it would be the only strictly correct method.” The problem with the right method is that nonsense is never allowed to get going; that is, as soon as something is said where no meaning has been given to a sign in the proposition, this is straight away brought to the attention of the speaker. It cannot satisfy and there would be no feeling of having been taught philosophy because the source of the philosophical concern that is motivating the speaker may not yet have been brought fully into focus, if at all. Patience and tolerance of nonsense is required in order to allow the concern that is motivating the speaker to express itself, as Wittgenstein attempts to do with his propositions concerning “philosophic matters” in the Tractatus. In contrast to the ‘right’ method, Wittgen- stein’s own allows nonsense to articulate itself in order for him to be able to identify and dispel the philosophical concern that is motivating the text. It entails letting nonsense be, in order for the metaphysical picture that is gripping the speaker to come into view. It entails giving the speaker freedom to assert ‘premises’ and deduce ‘conclusions’. But this ultimately is but a preliminary to the task of then showing the speaker that “he had giving no meaning to certain signs in his propositions.” The objective at this point is to recognise just how and in what way these propositions are nonsensical, whilst at the same time avoid- ing the possibility that this entails that there could be any conception of nonsense other than an austere one. As well as rejecting positivism, ‘resolutism’ also rejects what ineffabilism then wants to maintain, namely that a nonsensical proposition can nevertheless have an intelligible content that can be grasped in thought, but which is incapable of being expressed as a signiﬁcant proposition. What distinguishes ‘resolutism’ as a reading of the Tractatus is that it rejects any positivistic account of how propositional signs are to be judged nonsensical and any ineffabilistic account of what sense a nonsensical proposition may have.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 257 According to ‘weak resolutism’ the “grave errors” of the Tractatus are to be identified with an unacknowledged metaphysics concerning the essence of the logical structure of any representational language and of giving a complete analysis of this logical structure. Like, the strong version, it recognises that nonsense in the Tractatus is ultimately parasitic on a particular conception of sense, a conception that is explicated in terms of logical form. It regards this as an unacknowledged metaphysical commitment to the idea that there are hidden necessities and possibilities that determine the limits of the use of language. However, the weak version, unlike the more established readings of the Tractatus, does not think that this entailed any ineffable understanding of logical form. It believes that this can be illustrated by reference to the ‘say–show’ distinction in the Tractatus (see Part 1, above). The account that the weak version of resolutism gives of this ‘showing’ turns on the status of logical category distinctions, and in particular whether they can be ‘said’ using a significant proposition, or whether they can only be ‘shown’ using a concept script. The weak version believes that in the Tractatus Wittgenstein thought the latter, but, it should be stressed, that what is ‘shown’ are not ineffable truths about the nature of a ‘reality’ that obtains beyond the limits of language. What the ‘say–show’ distinction reveals is not the ineffability of nonsense, but perhaps ‘the ineffability of sense’, for what it is to make sense cannot in the end be fully made explicit by significant propositions; it requires a concept-script to make perspicuous the logical structure of language. What ‘resolutism’ regards as unintelligible, as opposed to being false, is any possibility of there being an understanding of how logical category distinctions are ‘shown’ in language that relies upon Realist or Anti-Realist notions of a form of ‘reality’ that determines or is determined by the form of language. In dismissing this, ‘resolutism’ does not then conclude that there is a problem concerning the exis- tence of an external world, or in talking about reality.What is being dismissed is only the Realist and Anti-Realist (mis)understanding of these terms. The form of language and the form of reality, discussed in the Tractatus, (the dropping of the scare quotes is intentional) are not two things that need to be brought into relationship to each other, but are less misleadingly said (say) to be one and the same thing considered from two different perspectives.

© Blackwell Publishing Ltd. 2003 258 Philosophical Investigations The strong version thinks that Wittgenstein is well aware of the metaphysics concerning logical form that nevertheless apparently underpins his criticism in the Tractatus of Realism and Anti-Realism. It regards this as the expression of yet another impulse towards metaphysics – of the persistent need to say something philosophical about how things are. It is the last rung of the ladder that must be climbed before the entire ladder of the Tractatus can be thrown away. ‘Strong resolutism’, not unlike ‘weak resolutism’, requires the reader to come to realise that nonsense in the Tractatus is ultimately parasitic on a particular conception of sense, a conception that is explicated in terms of logical form, and that if there is no such thing as the ineffability of nonsense, then there is no such thing as the effability of sense either.The purpose of the ‘say–show’ distinction is to call into question that a complete analysis can be given, for how sense is determined cannot in the end be said. Both the ineffability of nonsense and ultimately the effability of sense are called into question by the ‘say–show’ distinction. The propositions of the Tractatus prove to be both elucidatory and nonsense. But then nothing of the ‘Body’ or the ‘Frame’ remains to hang on to. The reader begins with the idea that limits need to be established to thinking and to the expression of thoughts, so that sense can be differentiated from nonsense. How such limits can be determined is then discussed, which involves in places a technical and detailed examination of the nature and application of logic. The outcome of this discussion is not just that the reader has a clearer understanding of what logic is and does, but that (more importantly) the very idea that there could be any such thing as meaning anything by ‘the limits to thinking’ is to be thrown away as nonsensical.What the reader is left with is the realisation that there is thinking going on as the propositions of the Tractatus are engaged with, but without thought in anything like the Fregean–Russellian sense. Logical analysis can neither fully capture nor fully specify just what it is to think. Thought and language cannot be ‘pinned down’ and the attempt to give a complete analysis breaks down; but not because it is impossible, for this suggests that there is something here that cannot be done, but because the very notion of given a complete analysis is unintelligible.There is nothing that amounts to anything in the notion of a vantage-point wherefrom we can survey thinking, where we can get beyond it and see it laid out neatly before us. (If you like: it always remains one step ahead of us, always in motion, always thwarting any attempt to circumscribe its bound-

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 259 aries.) The very attempt to do so, as the Tractatus, is intended to demonstrate, inevitably results in metaphysics, in ‘philosophical propositions’, in nonsense. However, none of this precludes a continuing role for logical analysis and for a concept-script.They could remain tools for helping make perspicuous how our language it actually is used;23 but they ‘cannot’ provide a complete analysis, establish ‘limits’ or provide ‘foundations’, for there is nothing that we are remotely likely tenably to regard as any such thing(s).This continuing role for logical analysis and for a concept script, if any, is therefore fundamentally therapeutic. This usefulness, however, is tempered by the possibility that logical analysis and a concept script, if misconceived, will become yet another expression of the impulse towards metaphysics. The propositions of the Tractatus then exploit this possibility in order to wean the reader away from its attraction. According to ‘resolutism’, it was never the aim of Wittgenstein in the Tractatus to give an account of how logical syntax permits or dis- allows possible combinations of words to make or fail to make sense. As we discussed in Part 1, it is speakers of language that ‘determine’ sense or nonsense, and only in contexts of signiﬁcant use. However, in overcoming the search for foundations that has preoccupied traditional forms of philosophical inquiry the Tractatus seemingly over- came philosophy itself.With the dispelling of the problems that give rise to traditional forms of philosophical inquiry there was no more need for philosophy as an ongoing deﬂationary and therapeutic activity. Compare this with what he writes in Section 133 of the Investigations: “[T]he clarity we are aiming at is indeed complete clarity. But this simply means that the philosophical problems should completely disappear. The real discovery is the one that makes me capable of stopping doing philosophy when I want to. – The one that gives philosophy peace, so that it is no longer tormented by questions which bring itself in question. – Instead, we now demonstrate a

23. For more detail on this, see Read’s “Logicism and Anti-Logicism are equally bankrupt and unnecessary”, in Haller and Puhl (eds.), Wittgenstein and the future of philosophy,[Proceedings of the Austrian Ludwig Wittgenstein Society (2001/2)], 380–8. Read here points out that, contrary to popular belief, Wittgenstein’s later writings and lectures on the philosophy of maths do not attempt to proclaim the uselessness or impossibility in general of a concept-script.

© Blackwell Publishing Ltd. 2003 260 Philosophical Investigations method, by examples; and the series of examples can be broken off. – Problems are solved (difficulties eliminated), not a single problem. There is not a philosophical method, though there are indeed methods, like different therapies.” The significance of this, for the strong version, is that it casts some light upon the nature of the continuities and discontinuities between the Tractatus and the Investigations, and the basis of Wittgenstein’s criticism of his former self. The clarity sought in the Tractatus was complete clarity, and that clarity required that philosophical problems should completely disappear. But the ‘real discovery’ is, we think, a recognition that the world we are returned to at the end of the Tractatus turns out to be a world in which the impulse towards metaphysics remains. It is also recognition of a legitimacy for philosophy. Peace comes to philosophy through allowing the continuing need for a deflationary and therapeutic philosophy in our lives; through accepting that there can almost certainly be no once-and-for-all dissolution to the problems of philosophy. What the Tractatus failed sufficiently to appreciate, according to the strong version, is not only that there are many different ways of misunderstanding ‘the’ logic of our language, and that each misunderstanding may require a different ‘solution’, but also that we can probably never totally free ourselves from the impulse towards metaphysics and so from the possibility of speaking nonsense, and therefore of the need of a therapeutic philosophy. This discovery enables us to stop doing philosophy when we want to and to start doing philosophy when we need to. It is a form of therapy that we may have a continual need for and the continuing possibility of engaging in. Significantly, it gives philosophy peace, (not us) as philosophy is no longer tormented by questions that bring itself (philosophy) into question. ‘Resolutism’ maintains that the Tractatus and the Investigations are both expressions of a therapeutic conception of philosophy. Wittgenstein’s later criticism of his former self concerns the ade- quacy of this expression. The weak version locates the grave errors that Wittgenstein later attributed to the Tractatus to an unacknowledged metaphysics concerning the essence of the logical structure of any representational language and of the need for a complete analysis of this logical structure. The strong version thinks that Wittgenstein recognised this account as an impulse towards meta-

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 261 physics – of the need to say something philosophical about how things are. It is seen as the last rung of the ladder that must be climbed before the entire ladder of the Tractatus can be thrown away. The strong version therefore locates the grave errors that Wittgen- stein later attributed to the Tractatus in how he had conceived of a therapeutic philosophy. The Tractatus can be seen as the expression of a single method, which, in turn, can be seen as a response to a single problem, generalised as a misunderstanding of ‘the’ logic of our language. In contrast, the Investigations demonstrates examples of speciﬁc misunderstandings of language in use, examples which are both person- and context- sensitive. The examples in the Investiga- tions are not only solutions to many problems; they also demonstrate many different methods. The diagnosis and cure of the Tractatus,in the end, failed to give philosophy peace; it continued to be “tormented by questions which bring itself in question”.

Part 3: ‘Resolutism’s’ Friendly Enemies: Vilhauer, Koethe and Emiliani

Vilhauer’s paper in the present issue of this journal is concerned with how Diamond’s austere conception of nonsense can accommodate Wittgenstein’s remark that the nonsense of the Tractatus is elucidatory. In particular, he thinks that there is a tension between Diamond’s account of nonsense and the explanation she gives of how nonsense elucidates.Vilhauer begins by identifying two key elements in Diamond’s view of nonsense: the ﬁrst commits her to the view that a propositional sign that is nonsensical does not have any logical segments; the second commits her to the view that a propositional sign that is nonsensical cannot be a logical segment within a proposition that is sensical. Vilhauer goes on to describes two essential stages in Diamond’s account of how nonsense elucidates: the ﬁrst is becoming conscious of the nonsensicality of the nonsense we utter; the second is continuing with our nonsensical utterances for the sake of elucidation, but being aware of their nonsensicality.The problem that Vilhauer thinks Diamond has is that the explanation she gives of how nonsense elucidates relies on there being logical relationships between propositional signs that are sensical and propositional signs that are nonsense of the kind that her account of nonsense rules out.

© Blackwell Publishing Ltd. 2003 262 Philosophical Investigations We acknowledge our deeper-than-deep indebtedness to Diamond’s groundbreaking work, but regard the early work from which Vilhauer quotes very much as prolegomena for a resolute reading, not as the reading itself. We believe that the account of ‘resolutism’ we give above clarifies and develops her seminal work in ways she might not accept but which overcome some of the criticisms that have been made of it. Diamond thinks that Wittgenstein inherits from Frege the distinction between saying and showing and the nonsensical character of elucidations. In Frege, these are used to grasp the indefinable nature of the primitive categorial distinctions that are required in order to be able to express genuine language by means of his Begriffsschrift. Diamond stresses the ‘Frame’ and regards the Tractatus as depicting and dispelling the metaphysical attractions of both Realism and Idealism through the imaginative use of nonsensical propositions. She believes that both Realism and Idealism prey upon a lack of clarity in our thinking and that a properly constructed Begriffsschrift can make perspicuous the inferential relations that obtain between propositional signs. Her conception of a Begriff- sschrift does seem to endorse the view that there is a single inferential order or underlying logical structure of language that it is the purpose of a Begriffsschrift to represent. However, what a Begriffsschrift shows are not ineffable features of ‘reality’, but the logical form of our language. There is of course very much in Diamond that we are in agreement with. However, in our opinion, Diamond’s account is in part susceptible to the criticisms that Vilhauer makes. This is because at times she inadequately distinguishes between the conclusion that a ‘philosophical proposition’ is nonsensical, and the process by which it is established that a propositional sign is a ‘philosophical proposition’; between a therapeutic conception of the continuing utility of logical analysis, and a metaphysical conception of what logical analysis is and can achieve; and therefore between those propositions of the Tractatus that are not about elucidating “philosophic matters” and so are perhaps sensical and those propositions that are about elucidating “philosophic matters” and are definitely nonsensical. Diamond’s schematic account veers at times toward being reductive; it does not sufficiently recognise the several different ‘levels’ at which the Tractatus operates, and to that extent it is over generalised. The key point here, that Vilhauer misses, is that neither the weak

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 263 nor the strong version of ‘resolutism’ need maintain, albeit for very different reasons, that all the propositions of the Tractatus are nonsensical, where nonsense is a technical term determined by the application of logical analysis to its own structure. The weak version (whether put forward in Diamond’s terms or not) would want to resist the conclusion that all the propositions of the Tractatus are nonsensical; the strong version would want to throw away the notion that thinking can be fully captured by and be fully specified by the application of logical analysis and so, ironically, resists the conclusion that anything could in the end be meant by magisterially declaring that the propositions of the Tractatus are one and all nonsensical. When 6.54 is put into this context, when the ‘propositions’ of the Tractatus are finally overcome, the reader becomes able at last to find sense and/or nonsense in its ‘propositions’ just in so far as the reader can see symbols in its signs, that is, is able to find a definite use for them. In both the weak and strong versions of ‘resolutism’ what comes after the throwing away of 6.54 is the recognition that the text remains there to be read and to be read on many different levels – levels that can be in tension with one another and that can call one another into question. ‘Resolutists’ believe that the text is so crafted that the reader can see-saw between sense and nonsense as its ‘propositions’ are engaged with and as contexts of significant use are explored. Further, ‘resolutism’ has no prescriptions whatsoever concerning the uses to which the ‘propositions’ of the Tractatus might be put. It invites others, as it is trying to do itself, to re-read the text again, but this time freed from the grip of the particular metaphysical pic- tures that the Tractatus has sought to dispel. However, if we have understood the author, then we will acknowledge that the impulse to metaphysics remains a continuing temptation and possibility, and consequently that whatever it is we find in the Tractatus may itself need to become the subject of the very deflationary and therapeutic activity that the book practices.24

24. This may have a paradoxical ring to it. Didn’t we maintain earlier that the strong version precisely applies destructively to all of the Tractatus, including the frame? But again, this is only an objection if we are taken to be ourselves providing a once-and-for- all key to how to read (or how to impugn) propositions of the Tractatus. We are not, and we are quite willing to engage in paradoxical thinking, in appearing to contradict ourselves, in order to attain philosophical insight. Our own remarks are intended as elucidations; it does not matter if they ‘contradict’ one another. (See also n. 26, below.)

© Blackwell Publishing Ltd. 2003 264 Philosophical Investigations Koethe, like Vilhauer, may (in his paper above) have correctly identified some unclarities in Diamond’s mode of expression in her revolutionary papers of nearly a generation ago now. But Koethe, (like Vilhauer, or perhaps still more so), has a very restricted conception of ‘resolutism’. Much like Mounce, what Koethe criticises is in fact only a form of positivism: namely, the view that all the propositions of the Tractatus are nonsense, where nonsense is a technical term determined by the application of logical analysis. What characterises both the weak and strong versions of resolutism is that each, in its own way, resists this conclusion. Much, therefore, of what Koethe says about the view (held by no ‘resolutist’) that he criticises, we agree with. The form of positivism that Koethe (wrongly) thinks the resolute reading is draws the wrong connections and obscures the more interesting affinities between Wittgenstein’s great early and late works. Whilst acknowledging that there is a therapeutic aspect to the Trac- tatus,Koethe thinks that the concentration on Wittgenstein’s ‘anti- philosophical’ tendencies overlooks what is a more central theme throughout his work: showing. (And he points out, as we do, that the ‘say–show’ distinction in the Tractatus cannot be used to support the ineffable truth of nonsensical propositions.) Both versions of resolutism make use of the ‘say–show’ distinction.The weak version maintains that what a Begriffsschrift supposedly shows are not ineffable features of ‘reality’, but the logical form of our language. This is part of the metaphysics of form that the weak version thinks that Wittgenstein later came to criticise in his early work. The strong version thinks that the purpose of the ‘say–show’ distinction is to undermine the belief that a complete logical analysis of language can be given, for how sense is determined ‘cannot’ in the end unmisleadingly be said to be said. The strong version therefore thinks that the ‘say–show’ distinction calls into question both the ineffability of nonsense and the effability of sense. In opposition to Koethe (and Mounce): neither the weak nor the strong version of ‘resolutism’ are likely to see the notion of ‘showing’ ultimately as having much significance in the later work. But this is not to say that the philosophical problems that the ‘say–show’ distinction was a response to are not present in Wittgenstein’s later work, but that they are dispelled, and in a quite different way.

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 265 Emiliani’s paper above offers a very full description and criticism of what he refers to as the “standard interpretation of logical form”: SILF, which he identifies with the more established (‘old’) readings of the Tractatus. An important part of Emiliani’s exegesis is what he refers to as the ‘immediacy of semantic agreement’. This is the view that the connection between the proposition and the state of affairs which it represents must be ‘internal’ to them; that there is no other intermediary element that secures this connection. Conse- quently, the unity and connection of the proposition just is the unity and connection of the state of affairs it represents. Emiliani also criticises what he refers to as the nihilist view,25 which he identifies with the resolute reading of the Tractatus. According to Emiliani, both SILF and the nihilist view reject the immediacy of semantic agreement. We are even more sympathetic to Emiliani than to Vilhauer and Koethe. But we have significant points of disagreement with him, too.The first is that we do not think that ‘resolutism’ must reject the immediacy of semantic agreement! On the contrary, our account of both versions of ‘resolutism’ holds that the unity and connection between a proposition and the state of affairs that it represents is an ‘internal’ one, and so is immediate, requiring nothing else to secure it.This follows directly from Wittgenstein’s contextualism in the Trac- tatus about sense and meaning.26 If this much is granted, then the real difference between us and Emiliani comes down to how ‘metaphysics’ is to be understood within the context of the Tractatus.We think that Emiliani would

25. We regard this label as unhelpful, for reasons which we hope emerged in Part 2. 26. It may be that Emiliani fails to see that ‘resolutism’ is not necessarily committed to rejecting the immediacy of semantic agreement because he thinks that ‘resolutism’ is committed to rejecting all philosophical views, and that believing in the immediacy of semantic agreement amounts to holding a philosophical view. There is something right about this. In the weak version of ‘resolutism’, Wittgenstein has philosophical views that he later saw to embody a metaphysical commitment and which he criticised. In the strong version of ‘resolutism’ Wittgenstein is aware that these philosophical views embody metaphysical commitments and that their use is fraught with the danger of being taken to be saying something, when the purpose is to realise that nothing is being said after all. All the talk in the Tractatus about the formal properties of propositions is designed to move us along in our appreciation of what logical analysis is and can achieve. It is transitional; and hence our points about ‘internality’ and about ‘contextualism’ are themselves elucidations; they are in the end to be overcome (or ‘thrown away’).

© Blackwell Publishing Ltd. 2003 266 Philosophical Investigations agree with us that underpinning SILF, and emphatically rejected by Wittgenstein not as false but as unintelligible, are Anti-Realist/ positivistic and Realist/ineffabilistic accounts of just what can and cannot be said or shown about the nature of a ‘reality’ that lies beyond the limit of what can be understood or known through natural scientiﬁc investigation. We think that he would agree with us that Wittgenstein emphatically rejects as unintelligible any notion that the ‘say–show’ distinction can be used to support the view that nonsensical propositions (as opposed to sensical propositions – the propositions of ‘science’; or senseless propositions – the ‘propositions’ of logic) ‘show’ ineffable features of ‘reality’. We think that these are among the “philosophic matters” that Wittgenstein intended the reader to come to recognise as nonsensical, as the attempt to say something metaphysical. But what of the account of logical form that is used to elucidate the essential unity and connection between the proposition and the state of affairs it describes, which is an attempt to say what can only be shown about the ‘relationship’ between language and reality (no scare quotes around those two words, as no underlying commitment to positivism/Anti-Realism or ineffabilism/Realism)? Emiliani, like the weak version of resolutism, maintains that what a Begriffsschrift shows are not ineffable features of ‘reality’, but the logical form of our language; but unlike the weak version he characterises this as metaphysical. Perhaps this is where our distinction between the weak and strong ‘resolutisms’ might sep- arate out something that Emiliani blurs. The weak version of resolutism does indeed think that there is a metaphysics of form in the Tractatus,but that Wittgenstein did not recognise it as such at the time of writing the Tractatus, and consequently that he only came criticise it much later. The strong version of resolutism also thinks that there is a metaphysics of form ‘in’ the Tractatus,but that Wittgen- stein did recognise it as such at the time of writing the Tractatus, and even used the ‘say–show’ distinction along the way to call into question both the ineffability of nonsense and the effability of sense. Emiliani comments at the end of his paper as to how the appraisal of nonsensical propositions in the Tractatus [concerning “philosophic matters”] can elucidate in a way that other expressions of nonsense cannot, and how this results in the destabilisation and disintegration of the formal properties of language that are made perspicuous by a Begriffsschrift. It seems to us that Emiliani is very close to embracing

© Blackwell Publishing Ltd. 2003 Rupert Read and Rob Deans 267 the conclusion of the strong version of resolutism that in the end it is not that a complete logical analysis of language cannot be given, but that nothing could be meant by the giving of a complete logical analysis of language. According to ‘strong resolutism’, there is nothing that we need to remain silent about once we have understood the author of the Trac- tatus.We suggested earlier a couple of senses in which the resolute reading, and not ineffabilism, is the true friend of saying and showing. But it can be equally apposite to say that, when one truly philosophizes in Wittgenstein’s spirit, early-and-late, nothing gets said, and nothing gets shown either. Philosophy, both as traditionally practised and as conceived by Wittgenstein in the Tractatus, can neither establish the limits of thought qua thought, nor the limits of the expression of thoughts in language. Also, neither of them can provide full and final criteria – a complete analysis – that can be used to demarcate sense from nonsense. (The ‘can’s’ in the above two sentences are to be read as we have urged throughout – as indexing a would-be task that we can come to recognise as non-existent, for principled reasons.) At the end of the Tractatus,we are returned to the world and to our lives with each other, both of which find meaningful expression or not, as the case may be, in the language that we speak. In Parts 1 and 2 of this paper, we said nothing that indicated explicit dissatisfaction with the ‘weak version’ of the (project of the) resolute reading of Wittgenstein. But perhaps the truth is now evident to the reader: We, like Juliet Floyd, favour some version of the strong version of ‘resolutism’.27 ‘Strong resolutism’ holds that the Tractatus was imperfect, but that the ‘places’ where these imperfec- tions are actually to be found are not even all the ‘places’ wherein Conant and Diamond are happy to declare the Tractatus flawed. We hope that the debate over Wittgenstein’s Tractatus henceforth (I) shows some awareness of the significant differences between the weak and strong versions of ‘resolutism’, and (II) pays more attention to the importance (and rarity) of a resolutist approach to all of Wittgenstein’s philosophy.

27. The ‘strong’ version is basically what Warren Goldfarb has called ‘Jacobinism’, the ‘weak’ version (closer to Conant and Diamond) what he has called ‘Girondinism’.

© Blackwell Publishing Ltd. 2003 268 Philosophical Investigations For all their virtues, none of the four papers we have responded to in this short piece fulﬁll either desideratum. And the greatest prize of all this is not scholarly machinations over the Tractatus; it is a new understanding – an understanding – of Philosophical Investigations.28

Philosophy SOC UEA Norwich NR4 7TJ UK

28. Our overwhelming debt in the writing of this paper is to Alice Crary, who gen- erously made freely available to us an unpublished manuscript from which many of the ideas in Part 1 of this paper were culled. [Any errors in this paper are, of course, our own.]